JP2007044331A - Testing apparatus for apnea - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a testing apparatus for apnea capable of precisely classifying the apnea in a simple configuration. <P>SOLUTION: The testing apparatus for apnea comprises an apnea determining means 20 to calculate an oxygen saturated concentration SpO<SB>2</SB>based on red pulse wave data D<SB>R</SB>(t) and infrared pulse wave data D<SB>IR</SB>(t) obtained by a pulse wave sensor 1 and to determine an apnea duration period T<SB>E</SB>-T<SB>S</SB>on a lowering of the oxygen saturation SpO<SB>2</SB>, a composite pulse wave data generating means 30 to compose the red pulse wave data D<SB>R</SB>(t) and the infrared pulse wave data D<SB>IR</SB>(t) and to obtain composite pulse wave data D<SB>G</SB>(t) corresponding to pulse wave data of a third wavelength light, in which absorptivity of aerobic hemoglobin and that of anaerobic hemoglobin are substantially equal, a data extracting means 40 to obtain low-frequency data D(t) of a frequency band having an upper cutoff frequency lower than a heartbeat frequency and including a breath frequency F<SB>B</SB>from the composite pulse wave data, and an apnea classifying means 50 to classify the apnea by pulsation degree based on the data of the apnea duration period T<SB>E</SB>-T<SB>S</SB>in the low-frequency data D(t). <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an apnea test apparatus suitable for diagnosis of sleep apnea syndrome and the like, and more particularly, to an apnea test apparatus that can make a simple and highly accurate determination of apnea type.

  In Japanese Patent Laid-Open No. 2002-153432, pulse wave data obtained from a pulse wave sensor worn on the wrist is passed through a low-pass filter to cut unnecessary frequency components (noise of 3 Hz or more due to body movement or external light), Various diagnostic methods for determining the type of apnea (central apnea, obstructive apnea, mixed apnea) by performing various post-processing on the low frequency data are disclosed. For example, the determination is made from the pulse wave amplitude or the waveform length of the low frequency data.

Central apnea is a condition in which the abdominal wall or thorax movement is also stopped due to respiratory stop due to abnormalities in the brain center. Obstructive apnea is a respiratory stop caused by obstruction of the respiratory tract due to fat or tongue. Alternatively, thorax movement is still occurring, and mixed apnea refers to a state in which obstructive apnea occurs following central apnea.
JP 2002-153432 A

  The heart rate (pulse) frequency is about 1 Hz on an adult average, and the respiratory frequency is about 0.4 Hz on an adult average. In the invention according to the above publication, in the pulse wave data at a low frequency of 3 Hz or less, the heart rate frequency component is prominently present as the main pulse wave regardless of the type of apnea, and the lower respiratory frequency component is the heart rate frequency component. Attached as fluctuations. Although there are various metric expressions for evaluating this fluctuation, as shown in the above publication, the correlation between the intensity of the respiratory frequency component and the degree of fluctuation cannot be ensured due to the inclusion of unnecessary frequency components other than the respiratory frequency. In the invention according to the above publication, the light emitting element used for the pulse wave sensor is single, and it is not described whether or not it is a monochromatic light source. However, depending on the emission wavelength, aerobic (oxidized) hemorrobin Since the extinction coefficient of oxygen and the extinction coefficient of oxygen-free (reduced) hemoglobin do not coincide with each other, a component reflecting the fluctuation of the oxygen saturation concentration is mixed in the pulse wave data, and the fluctuation itself due to the respiratory frequency component is disturbed. For this reason, in the method of determining the type of apnea from the fluctuation of the pulse wave data, the determination accuracy is poor and the reliability is poor.

  Therefore, in view of the above problems, an object of the present invention is to provide an apnea test apparatus capable of accurately determining the type of apnea while having a simple configuration.

  By the way, the breathing frequency is the ventilation (breathing) frequency by the mouth or nose due to the abdominal wall or thorax movement, but the mouth airflow or nasal airflow is stopped, the oxygen saturation concentration is lowered, and it is determined as the apnea time. Even in the case of obstructive apnea or mixed apnea, since there is still an abdominal wall or thorax movement, the influence also affects the arterial blood layer that is the detection target of the pulse wave sensor. The heartbeat motion causes a heartbeat pulsation of hemoglobin flow that flows inside the blood vessel, but the respiratory motion, that is, the abdominal wall or thorax motion distorts from the outside of the blood vessel, causing the pulsation of hemoglobin flow. The inventors of the present invention have observed that the directing of the periodic motion of the abdominal wall or thorax as pulsation using a pulse wave sensor that measures the absorption of hemoglobin is the direction of solving the problem.

  In order to solve the above problems, a first apnea examination apparatus according to the present invention includes a pulse wave data collection means 10, an apnea determination means 20, and third pulse wave (synthetic pulse wave) data generation as shown in FIG. Means 30, data extraction means 40 and apnea type determination means 50.

The pulse data collection means 10 includes the pulse wave sensor 1 and the like, and irradiates the arterial blood layer TG with the first wavelength light (for example, the red light R) and the second wavelength light (for example, the infrared light IR). Absorbed light (reflected light or transmitted light) is received, and time-series first (red) pulse data D _R (t) and time-series second (infrared) pulse data D _IR (t) are obtained. obtain. The apnea determination means 20 calculates a time-series oxygen saturation concentration based on the first (red) pulse data D _R (t) and the second (infrared) pulse data D _IR (t), and this oxygen saturation The apnea period is defined by the decrease in concentration. The third pulse wave (synthetic pulse wave) data generating means 30 combines the first (red) pulse group data D _R (t) and the second (infrared) pulse group data D _IR (t) by weighting or the like, Third pulse group (synthetic pulse wave) data D _G corresponding to time-series pulse group data of third wavelength light (for example, green light) in which the extinction coefficient of aerobic hemoglobin and that of anaerobic hemoglobin are substantially equal (T) is obtained. The data extraction means 40 obtains low frequency data D (t) in a frequency band having a high-frequency cutoff frequency lower than the heartbeat frequency and including the respiratory frequency from the third pulse group (synthetic pulse wave) data D _G (t). The apnea type determination means 50 determines the apnea type from the degree of pulsation for the apnea period data over the apnea period in the low frequency data D (t).

In the apnea test apparatus having such a configuration, the third pulse wave (synthetic pulse wave) data generating means 30 generates the third wavelength light having substantially the same absorptive coefficient of aerobic hemoglobin and absorptive hemoglobin. Since the time-series third pulse wave (synthetic pulse wave) data D _G (t) is obtained, the third pulse wave (synthetic pulse wave) data D _G (t ), The component reflecting it is not mixed in, and shows a transition according to the hemoglobin flow rate, so the pulsation reflecting the periodic motion of the abdominal wall or thorax is conspicuous. Moreover, the data extraction means 40 uses the third pulse wave (synthetic pulse wave) data D _G (t) to generate low frequency data D (t ), The heart rate frequency component is not mixed into the low frequency data D (t), so the respiratory frequency (abdominal wall or thorax motion frequency) component is more conspicuous. Appear. Moreover, it becomes flat because there is no pulsation in the case of central apnea, and further, pulsation appears after the flat in the case of mixed apnea.

For this reason, the determination accuracy by the apnea type determination means 50 is improved, and the reliability of the determination result of the apnea type is improved. In addition, since only the pulse wave sensor attached to the wrist or the like can collect data necessary for determining the apnea type, it can be examined without giving the subject a sense of restraint during sleep, and the apparatus configuration can be simplified. . Further, the third pulse wave (synthetic pulse wave) data D _G (t) is the first pulse wave data D _R (t) and the second pulse wave data for calculating the oxygen saturation concentration used for defining the apnea period. Since D _IR (t) is synthesized, the third pulse wave light source or the like can be omitted in the pulse wave sensor, the complication of the configuration of the pulse data collection means 10 can be avoided, and the cost can be reduced.

The second apnea examination apparatus according to the present invention does not include the third pulse wave (synthetic pulse wave) data generation means 30 in the first apnea examination apparatus, as shown in FIG. The first-wavelength light, the second-wavelength light, and the third-wavelength light (for example, green light G) in which the extinction coefficient of aerobic hemoglobin is substantially equal to the extinction coefficient of anaerobic hemoglobin are applied to the arterial blood layer TG. Non-absorbed light is received, and time-series first pulse data D _R (t), time-series second pulse data D _IR (t), and time-series third pulse wave data D _G (t) The pulse data collecting means 10 'for obtaining the above is provided.

  In such a configuration, the pulse wave data collecting means 10 ′ to which a third pulse wave light source and the like are added is obtained, but the operational effect is the same as that of the first apnea test apparatus.

  As said apnea determination means 20, it is desirable to calculate the apnea start time as a time point a predetermined time (for example, 20 seconds) before the start of the apnea period. This is because the decrease of the oxygen saturation concentration is assumed to occur with an average of 20 seconds from the apnea start time. By recording the apnea start time and apnea period in the memory 60, and reading and displaying them after waking up, the subject himself / herself can know the presence / absence of apnea, the number of times, apnea start time, apnea time, and the like.

  Although the data extraction means 40 has a function as a band pass filter, it is desirable that the frequency band has a low cut-off frequency higher than the frequency near 0 Hz. Extremely low frequency noise such as swell components of blood vessels due to extraneous light, body motion, and pulse can be cut, and unnecessary fluctuations in the low frequency data D (t) can be suppressed. Moreover, as a frequency band, it is desirable that a predetermined respiratory frequency is defined as an intermediate frequency between the high-frequency cutoff frequency and the low-frequency cutoff frequency. This is because the subject's respiratory frequency slightly increases or decreases. For example, the high-frequency cutoff frequency may be a frequency that is 0.1 Hz higher than a predetermined respiratory frequency, and the low-frequency cutoff frequency may be a frequency that is 0.1 Hz lower than the predetermined respiratory frequency.

However, if the predetermined respiration frequency is used, the frequency of mismatch due to individual differences and time zones increases. Therefore, every time the apnea determination means 20 defines the apnea period, the third pulse wave (synthetic pulse wave) It is desirable to provide a breathing frequency detecting means 42 for detecting the breathing frequency after frequency analysis of the data D _G (t). Since the breathing frequency at the time when apnea occurs peculiar to the subject, pulsation due to the breathing frequency peculiar to the subject can be emphasized in the low frequency data D (t).

By the way, the provision of the respiratory frequency detection means 42 requires the provision of Fourier transform means for frequency analysis of the third pulse wave data D _G (t). Since the data extraction unit 40 performs a bandpass function, when this Fourier transform unit is used, the data extraction unit 40 performs fast Fourier transform processing on the third pulse wave data D _G (t) including the apnea period. a FFT unit 41 to obtain the frequency characteristic data a (f) is subjected to a breathing frequency detection means 42 for detecting a respiration frequency F _B on the frequency characteristic data a (f), of the frequency characteristic data a (f) A configuration having inverse FFT means 43 that performs high-speed inverse Fourier transform processing on the frequency band and obtains the low-frequency data D (t) can be employed. The data extracting unit 40 combines the bandpass function of the frequency band is defined by using a detection and respiratory frequency F _B of the subject-specific respiratory frequency F _B, it can be realized to shorten the simple and the processing time of the construction .

Note that the breathing frequency detection means 42, it is desirable to set the frequency of the spectrum and respiratory frequency _{F B} in the range of 0.1Hz~0.6Hz of the frequency characteristic data A (f).

  From the degree of pulsation of the low-frequency data D (t), for example, the number of pulse waves can be counted to directly determine the apnea type. However, since the pulse wave amplitude may be small, the threshold setting is delicate, but a small amplitude pulse wave that reflects the respiratory frequency cannot be ignored. Therefore, the apnea type determination means 50 preferably includes a waveform deformation processing means for exaggerating the pulsation mode of the apnea period data over the apnea period of the low frequency data D (t) based on the respiration frequency. . Since the periodicity of the pulse wave is exaggerated and deformed in association with the already obtained subject-specific respiratory frequency information, the degree of pulsation can be manifested, and the apnea type determination accuracy is increased.

As an example of the waveform deformation processing unit, the apnea type determination unit 50 may employ a configuration including a normalization processing unit 51 that creates normalized data R (t) in the range 0 to 1. Assuming that the half cycle of the respiratory frequency is ΔT, the apnea period data at time t is D _E-S (t), and the apnea data at time (t + ΔT) is D _E-S (t + ΔT), the normalized data R ( t) is D _E−S (t + ΔT) / D _E−S (t) when D _E−S (t) ≦ D _E−S (t + ΔT), and D _E−S (t)> D _{When E−S} (t + ΔT), D _E−S (t) / D _E−S (t + ΔT). Even if the amplitude of the pulse wave synchronized with the respiration frequency in the low frequency data D (t) is small, the interval between the apex and the bottom is the half cycle ΔT of the respiration cycle, so the low frequency data D (t) By autocorrelating with the period ΔT, if there is a pulse wave in the low frequency data D (t), the portion of the pulse wave corresponding to the respiratory frequency can be emphasized in the apnea period data D _E-S (t). Apnea determination is easy. In addition, since the ratio is calculated, when the low frequency data D (t) is flat, fluctuations can be suppressed in the apnea period data D _E-S (t), and central apnea can be easily determined. Become.

As a quantification method for the type determination, a histogram generation unit 52 that generates a histogram in which the section i subdividing the range 0 to 1 and the number of data by section n (i) belonging to this section correspond to each other; It is desirable to provide a central degree calculation means 53 for calculating a ratio (M / N) of the normalized data R (t) to the total data number N of the section-specific data number M belonging to the section closest to 1 in the histogram. . Then, the ratio (M / N) is determined when exceeding compares whether more than a first threshold value TH ₁ near 1 and central apnea, the ratio (M / N) a second threshold value TH close to 0 when ₂ smaller than or less or not than determines the closed apnea, the ratio (M / N) is compared whether the second threshold value TH ₂ or more in the first threshold value TH ₁ below both thresholds In the meantime, it is desirable to provide a comparison means 54 for determining mixed apnea.

  By recording this type determination result in the memory 60 and reading and displaying it after awakening, the subject can also know the result.

  In the present invention, the third pulse wave data generation means obtains the third pulse wave data of the time series of the third wavelength light, in which the extinction coefficient of aerobic hemoglobin and the extinction coefficient of anaerobic hemoglobin are substantially equal. Since the third pulse wave data shows a transition according to the hemoglobin flow rate, the pulsation reflecting the periodic motion of the abdominal wall or thorax stands out, and the heart rate is cut by the data extraction means. Since the abdominal wall or thorax motion frequency components further stand out in the frequency data, the apnea type determination accuracy is increased and the reliability is improved. In addition, since only the pulse wave sensor attached to the wrist or the like can collect data necessary for determining the apnea type, it can be examined without giving the subject a sense of restraint during sleep, and the apparatus configuration can be simplified. .

  Next, embodiments of the present invention will be described with reference to the accompanying drawings.

  FIG. 1 is a functional block diagram showing an apnea test apparatus according to a first embodiment of the present invention, FIG. 2 is a flowchart showing an operation flow from data sampling to apnea type determination in the apparatus, and FIG. FIG. 4 is a flowchart showing an apnea type determination routine in the apparatus.

  As shown in FIG. 1, the apnea examination apparatus of this example includes a pulse wave data collection means 10, an apnea determination means 20, a synthetic pulse wave data generation means 30, a data extraction means 40, an apnea type determination means 50, and a memory. 60 and so on. In this example, the apnea determination means 20, the synthetic pulse wave data generation means 30, the data extraction means 40, the apnea type determination means 50, and the memory 60 are constituted by a microcomputer.

The pulse data collection means 10 includes a wristwatch-type pulse wave sensor 1 attached to a wrist or the like, and an encoder / decoder circuit 2 that controls the sensor and receives a sensor output. The pulse wave sensor 1 includes a pulse width modulation (PWM) circuit 1a for alternately emitting light of a red light emitting part (red light emitting diode) 1b having a wavelength of 660 nm and an infrared light emitting part (infrared light emitting diode) 1c having a wavelength of 905 nm. The red light R and the infrared light IR alternately irradiate the arterial blood layer TG such as the wrist, and have a single light receiving portion (photodiode) 1e that alternately receives the reflected light therefrom. The encoder of the circuit 2 controls the pulse width modulation (PWM) circuit 1a, and the decoder of the circuit 2 performs A / D conversion on the sensor output from the light receiving unit 1e, and then distributes at a predetermined sampling period in synchronization with the encoder. A series of red pulse wave data D _R (t) and time series of infrared pulse wave data D _IR (t) are generated.

The apnea determination means 20 calculates a time-series oxygen saturation concentration (SpO ₂ ) based on the red pulse data D _R (t) and the infrared pulse data D _IR (t), and reduces the oxygen saturation concentration. The apnea duration (T _E -T _S ) is defined as an opportunity. Moreover, the apnea determination means 20 calculates the apnea start time (T _S- 20) as the apnea start time (T _S- 20), and calculates a time point 20 seconds before the start (T _S ) of the apnea period (T _E- T _S ) The start time (T _S −20) and the apnea duration (T _E −T _S ) are stored in the memory 60. The synthetic pulse wave data generation means 30 synthesizes the red pulse data D _R (t) and the infrared pulse data D _IR (t) by weighting and the like, and the absorption coefficient of aerobic (oxidized) hemoglobin (HbO ₂ ) Synthetic pulse group data D _G (t) corresponding to the time-series pulse group data of the green pulse wave, which is substantially equal to the extinction coefficient of oxygen-free (reduced) hemoglobin (Hb), is obtained.

The data extraction means 40 obtains low-frequency data D (t) in a frequency band having a high-frequency cutoff frequency lower than the heartbeat frequency and including the respiratory frequency from the synthetic pulse group data D _G (t), and apnea FFT means 41 that obtains frequency characteristic data A (f) by performing fast Fourier transform on the synthesized pulse wave data D _G (t) including the duration (T _E −T _S ), and the frequency characteristic data A (f) and respiration frequency detection means 42 for detecting a respiration frequency F _B above, the inverse FFT unit 43 to obtain a low-frequency data D (t) subjected to inverse fast Fourier transform processing on the frequency band of the frequency characteristic data a (f) have.

The apnea type determination means 50 determines the apnea type from the degree of pulsation for apnea period data over the apnea duration (T _E -T _S ) of the low frequency data D (t). A processing unit 51, a histogram generation unit 52, a central degree calculation unit 53, and a comparison unit 54. The normalization processing means 51 generates normalization data R (t) in the range 0 to 1, the half cycle of the respiratory frequency is ΔT, the apnea data at the time t is D _E−S (t), and the time (t + ΔT) apnea duration data) when _{D E-S (t + ΔT} ), the normalized data R (t) _may, when _{D E-S (t) ≦} D E-S of _{(t + ΔT) D E-} S (t + ΔT ) / D _E−S (t), where D _E−S (t)> D _E−S (t + ΔT), D _E−S (t) / D _E−S (t + ΔT). The histogram generation means 52 generates a histogram in which the section i subdividing the range 0 to 1 corresponds to the number of data by section n (i) belonging to the section i. The centrality degree calculating means 53 calculates the ratio (M / N) of the number of data M by section belonging to the section closest to 1 in the histogram to the total number of data N of the normalized data R (t). The comparison means 54 determines whether or not the ratio (M / N) is close to the first threshold TH ₁ that is close to 1 and determines that the ratio is (M / N) close to 0. 2 when the threshold TH small compared or less or not than ₂ determines the closing apnea, compare whether the ratio (M / N) is the second threshold value TH ₂ or more in the first threshold value TH ₁ or less If it is between both threshold values, it is determined as mixed apnea and the determination result is stored in the memory 60.

Next, the operation of the apnea test apparatus of this example will be described in detail with reference to FIG. The pulse wave sensor 1 alternately irradiates the arterial blood layer TG with red light R and infrared light IR, and in step ST1, the encoder / decoder circuit 2 samples the sensor output at 16 Hz, and the time series shown in FIG. Red pulse wave data D _R (t) and time-series infrared pulse data D _IR (t) shown in FIG. 5B are generated. In step ST2, the apnea determination means 20 calculates the oxygen saturation concentration SpO ₂ (t). The calculation routine as shown in FIG. 3, in step S1, the amplitude ratio ΔAC the AC component AC _R and alternating current component _{AC IR} the infrared pulse wave data _D IR (t) of the red pulse wave data _D R (t) ( = calculates the _AC IR / _{AC R),} then in step S2, the level of the direct current component _{DC IR} DC component DC _R and infrared pulse wave data _D IR (t) of the red pulse wave data _D R (t) After calculating the ratio ΔDC (= DC _IR / DC _R ), in step S3, the oxygen saturation concentration SpO ₂ (t) = a × Log 10 (ΔAC × ΔDC) + b is calculated with the calibration coefficients a and b.

Next, in step ST3, after storing the oxygen saturation concentration SpO ₂ (t) of the number of data for the past 300 seconds shown in FIG. 6A, in step ST4, the synthetic pulse wave generating means 30 uses the weighting coefficient ΔI. The red pulse data D _R (t) and the infrared pulse data D _IR (t) were synthesized ({ΔI · D _R (t) + D _IR (t)} / 2) and converted into a green light pulse wave. Synthetic pulse wave data D _G (t) is obtained, and in step ST5, the synthetic pulse wave data D _G (t) for the past 300 seconds shown in FIG. 6B is stored.

In step ST6, the apnea determination means 20 searches the data of the oxygen saturation concentration SpO ₂ (t) for the past 300 seconds, and determines whether there is data that has decreased to 90% or less with time. When it does not decrease, the process returns to step ST1, but when it decreases, in step ST7, the apnea determination means 20 determines whether or not the decrease starts to recover. Returning to step ST1 if not started to recover, in step ST8 if began to recover, apnea determining means 20, oxygen saturation SpO 2 _(t) is calculated when T _E which began to recover, step In ST9, a time T _{S at} which the oxygen saturation concentration SpO ₂ (t) starts to decrease is calculated, and in step 10, it is determined whether or not the apnea duration (T _E −T _S ) is 10 seconds or more. If it is less than 10 seconds, the process returns to step ST1, but if it is 10 seconds or more, an apnea determination result is obtained, and in step 11, the apnea type determination routine shown in FIG. Is executed.

First, in step T1, the FFT means 41 of the data extraction means 40 performs fast Fourier transform (frequency analysis) processing on the synthetic pulse data D _G (t), and the frequency characteristic data A (f shown in FIG. ). In this A (f) characteristic, the frequency of the spectrum near 1 Hz is the heartbeat (pulse) frequency, and the frequency near 0 Hz is mainly a noise component including a body motion component and a swell component of the blood vessel due to the pulse, 0.1 Hz to 0.6 Hz. The spectrum that appears in the range is the component of the respiratory frequency F _B , and the body motion component also appears between F _{B and 1} Hz. Therefore, since the breathing frequency detection means 42 is to detect and breathing frequency frequency _{F B} of the spectrum is in the range of 0.1Hz~0.6Hz of the frequency characteristic data A (f) shown in FIG. 7 (a), it is possible to detect the respiration frequency F _B at which an apnea condition of the subject has occurred. In step T2, the inverse FFT means 43 performs high-speed operation in the frequency band of the low frequency cutoff frequency (F _B −0.1) Hz to the high frequency cutoff frequency (F _B +0.1) Hz in the frequency characteristic data A (f). Inverse Fourier transform processing is performed to generate low frequency data (breathing curve) D (t) shown in FIG. Respiratory frequency F _B is the center frequency of the frequency band.

In step T3, the normalization processing means 51 of the apnea type determination means 50 selects data over the apnea period (T _E −T _S ) of the low frequency data (respiration curve) D (t), and ranges from 0 to 0. 1 normalization data R (t) is generated. When the half cycle 1 / 2F _B of the respiratory frequency F _B is ΔT, the apnea data at the time t is D _E-S (t), and the apnea data at the time (t + ΔT) is D _E-S (t + ΔT), Normalized data R (t) is as follows, as shown in FIG.

When D _E−S (t) = B3 ≦ D _E−S (t + ΔT) = B4, D _E−S (t) / D _E−S (t + ΔT) = B3 / B4 = R2, and D _{E When −S} (t) = B1> D _E−S (t + Δt) = B2, D _E−S (t + ΔT) / D _E−S (t) = B2 / B1 = R1.

  Next, at step T4, as shown in FIG. 8 (b), the histogram generating means 52 corresponds to the section i subdivided between the above range 0 to 1 and the number of data by section n (i) belonging to this section i. A histogram is generated. Further, the centrality degree calculating means 53 normalizes the normalized data R (the number of data by section M belonging to the section closest to 1 (1 ≧ R (t) ≧ 0.995) out of the number of data by section n (i). The ratio (100 × M / N)% of the total data number N of t) is calculated.

In step T5, the comparison means 54 determines whether or not the ratio (100 × M / N) exceeds the first threshold value TH ₁ (80%), and determines that the ratio is (100). When xM / N) is smaller than the second threshold TH ₂ (40%), it is determined that the type is obstructive apnea, and the ratio (100 × M / N) is the first threshold TH _1. comparing whether the second threshold value TH ₂ or below is determined as mixed apnea when between both thresholds.

In step T 6, the apnea type determination result, apnea start time (T _S −20), and apnea duration (T _E −T _S ) are stored in the memory 60.

FIG. 9A is a graph showing the transition of apnea period data D _E-S (t) of central apnea, and FIG. 9B is obtained from the apnea period data D _E-S (t). FIG. 9C is a graph showing the transition of the normalized data R (t). FIG. 9C shows the number of data by section n (i) obtained from the normalized data R (t). The transition of the apnea period data D _E-S (t) is flat. In the normalization, a pulse wave with a small amplitude is also exaggerated. Therefore, a pulsation appears a little in the normalized data R (t), but the ratio (100 × M / N) is equal to the first threshold value TH ₁ (80%). Since it has exceeded, it is determined as central apnea.

FIG. 10A is a graph showing the transition of the apnea period data D _E-S (t) of the obstructive apnea, and FIG. 10B is obtained from the apnea period data D _E-S (t). FIG. 10C is a graph showing the transition of the normalized data R (t), and FIG. 10C shows the number of data by section n (i) obtained from the normalized data R (t). Although pulsation appears throughout the apnea period data D _E-S (t), the pulsation is exaggerated in the normalization, and therefore the pulsation appears remarkably in the normalized data R (t), and the ratio (100 Since × M / N) is less than the second threshold TH ₂ (40%), it is determined as obstructive apnea.

FIG. 11A is a graph showing the transition of apnea period data D _E-S (t) of mixed apnea, and FIG. 11B is obtained from the apnea period data D _E-S (t). FIG. 11C is a graph showing the transition of the normalized data R (t). FIG. 11C shows the number of data by section n (i) obtained from the normalized data R (t). Since the apnea period data D _E-S (t) changes from flat to pulsation, the aspect is exaggerated in the normalized data R (t). Since the ratio (100 × M / N) is between the second threshold value TH ₂ (40%) and the first threshold value TH ₁ (80%), it is determined as mixed apnea.

As described above, since the synthetic pulse wave data D _G (t) corresponding to the time series of green light in which the extinction coefficient of aerobic hemoglobin and the extinction coefficient of anaerobic hemoglobin are substantially equal, the oxygen saturation concentration is obtained. Even if SpO ₂ fluctuates over time, the component reflecting the same is not included in the composite pulse wave data D _G (t), and shows a transition according to the hemoglobin flow rate. Therefore, the pulsation reflecting the periodic motion of the abdominal wall or thorax stands out in the synthetic pulse wave data D _G (t). Moreover, from the synthetic pulse wave data D _G (t), a frequency band (F _B −0.1 to F) having a high cut-off frequency (F _B +1) lower than the heartbeat frequency (about 1 Hz) and including the respiration frequency F _{B. Since} the low frequency data D (t) of _B + 0.1) is obtained, the heart rate frequency component is not mixed in the low frequency data D (t), so that the respiratory frequency (abdominal wall or thorax motion frequency) component further stands out. It appears clearly as pulsation during obstructive apnea. Moreover, it becomes flat because there is no pulsation in the case of central apnea, and further, pulsation appears after the flat in the case of mixed apnea. For this reason, the determination accuracy by the apnea type determination means 50 is improved, and the reliability of the determination result of the apnea type is improved.

In addition, since only the pulse wave sensor 1 attached to the wrist or the like can collect data necessary for determining the type of apnea type, it can be examined without giving a sense of restraint to the subject during sleep, and the apparatus configuration is simplified. it can. Further, the synthetic pulse wave data D _G (t) is obtained by using the first pulse group data D _R (t) and the second pulse group data D _IR (t) for calculating the oxygen saturation concentration used for defining the apnea period. Since they are synthesized, the third pulse wave light source and the like in the pulse wave sensor 1 can be omitted, the complication of the configuration of the pulse wave data collection means 10 can be avoided, and the cost can be reduced.

  FIG. 12 is a functional block diagram showing an apnea test apparatus according to Embodiment 2 of the present invention. The same parts as those shown in FIG.

As shown in FIG. 12, the apnea test apparatus of this example does not include the synthetic pulse wave data generation unit 30 in the apnea test apparatus shown in FIG. 1. Instead, the red light R, infrared light IR, and The green blood G having substantially the same extinction coefficient of aerobic hemoglobin and the extinction coefficient of anaerobic hemoglobin is irradiated to the arterial blood layer TG, and the non-absorbed light from each is received, and time-series red pulse data D _R ( t), pulse group data collecting means 10 'for obtaining time series infrared pulse group data D _IR (t) and time series green pulse wave data D _G (t) is provided. The pulse wave sensor 1 is additionally provided with a green light emitting unit 1c having a wavelength of 540 nm.

FIG. 13 is a flowchart showing the flow of operations from data sampling to determination of apnea type in this example. The same parts as those shown in FIG. The pulse wave sensor 1 cyclically irradiates the arterial blood layer TG with red light R, infrared light IR, and green light G, and in step ST1 ′, the encoder / decoder circuit 2 samples the sensor output at 16 Hz, Red pulse wave data D _R (t), time-series infrared pulse wave data D _IR (t), and time-series green pulse wave data D _G (t) are generated. In step ST5 ′, the green pulse wave data D _G (t) for the past 300 seconds is stored.

FIG. 14 is a flowchart showing an apnea type determination routine in this example. The same parts as those shown in FIG. In steps T1 ′ and T2 ′, the data D _G (t) is substantially equal only by replacing the combined pulse wave of steps T1 and T2 of FIG. 4 with the green pulse wave.

  In this example, the pulsation data collecting means 10 'is added with the green light emitting unit 1c, but the effect is the same as in the case of the apnea test apparatus shown in FIG.

It is a functional block diagram which shows the apnea test apparatus which concerns on Example 1 of this invention. It is a flowchart which shows the flow of operation | movement from data sampling to the determination of apnea type | mold in the same apparatus. It is a flowchart which shows the measurement routine of oxygen saturation concentration in the same apparatus. It is a flowchart which shows the determination routine of apnea type | mold in the same apparatus. In the same apparatus, (a) is a graph showing the transition of red pulse wave data, and (b) is a graph showing the transition of infrared pulse wave data in the apparatus. (A) is a graph which shows transition of oxygen saturation concentration in the apparatus, (b) is a graph which shows transition of synthetic pulse wave data in the apparatus. (A) is a spectrum diagram showing frequency characteristic data in the apparatus, and (b) is a graph showing synthetic pulse wave data and low frequency data in the apparatus. (A) is explanatory drawing explaining the aspect of the normalization process in the apparatus, (b) is a histogram produced | generated in the apparatus. (A) is a graph showing the transition of apnea data of central apnea, (b) is a graph showing the transition of normalized data obtained from the apnea data, (c) is from the normalized data It is the obtained histogram. (A) is a graph showing transition of apnea period data of obstructive apnea, (b) is a graph showing transition of normalized data obtained from the apnea period data, (c) is based on the normalized data It is the obtained histogram. (A) is a graph showing the transition of apnea period data of mixed apnea, (b) is a graph showing the transition of normalized data obtained from the apnea period data, (c) is from the normalized data It is the obtained histogram. It is a functional block diagram which shows the apnea test apparatus which concerns on Example 2 of this invention. It is a flowchart which shows the flow of operation | movement from data sampling to the determination of apnea type | mold in the same apparatus. It is a flowchart which shows the determination routine of apnea type | mold in the same apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1,1 '... Pulse wave sensor 1a ... Pulse width modulation (PWM) circuit 1b ... Red light emission part 1c ... Infrared light emission part 1d ... Green light emission part 1e ... Light reception part 2 ... Encoder / decoder circuit 10, 10' ... Pulse group Data collection means 20 ... apnea determination means 30 ... synthetic pulse wave data generation means 40 ... data extraction means 41 ... FFT means 42 ... respiration frequency detection means 43 ... inverse FFT means 50 ... apnea type determination means 51 ... normalization processing means 52 ... histogram generating means 53 ... central type calculation means 54 ... comparing unit 60 ... memory R ... red light IR ... infrared light G ... green light TG ... arterial layer _D R (t) ... red pulse wave data _D IR (t ) Infrared pulse wave data D _G (t) Synthetic (green) pulse wave data SpO ₂ Oxygen saturation concentration (T _E -T _S ) Apnea duration (T _S -20) Apnea start time A (F) Frequency characteristic data F B _... breathing frequency _{D E-S} (t) ... apnea duration data D (t) ... low-frequency data (breathing curve)
R (t) ... Normalized data n (i) ... Number of data by section TH ₁ ... First threshold TH ₂ ... Second threshold

Claims (13)

Pulse wave data collection means for irradiating the arterial blood layer with the first wavelength light and the second wavelength light and receiving non-absorbed light therefrom to obtain time-series first pulse data and time-series second pulse data. And an apnea determining means for calculating a time-series oxygen saturation concentration based on the first and second pulse data and defining an apnea period based on a decrease in the oxygen saturation concentration; and the first and second pulses The third pulse wave is obtained by synthesizing the group data and obtaining third pulse group data corresponding to the time series pulse data of the third wavelength light in which the extinction coefficient of the aerobic hemoglobin and the extinction coefficient of the anaerobic hemoglobin are substantially equal. Data generation means, data extraction means for obtaining low-frequency data in a frequency band having a high-frequency cutoff frequency lower than the heartbeat frequency and including a respiratory frequency from the third pulse data, and the apnea period of the low-frequency data Apnea period data For in apnea inspection apparatus characterized by comprising and an apnea type determining means for determining an apnea type pulsation degree. First wavelength light, second wavelength light, and third wavelength light with substantially the same absorption coefficient of aerobic hemoglobin and anaerobic hemoglobin are applied to the arterial blood layer to receive non-absorbed light therefrom. , A pulse series data collecting means for obtaining time-series first pulse data, time-series second pulse data, and time-series third pulse wave data, and time series based on the first and second pulse data And the apnea determination means for defining the apnea period based on the decrease in the oxygen saturation concentration, and the low frequency data in the frequency band including the respiration frequency having a high cut-off frequency lower than the heartbeat frequency. Data extraction means obtained from the third pulse group data, and apnea type determination means for determining the apnea type from the degree of pulsation of the apnea period data of the low frequency data over the apnea period. thing Apnea inspection apparatus for the butterflies. 3. The apnea test apparatus according to claim 1, wherein the apnea determination unit calculates a time point a predetermined time before the start of the apnea period as an apnea start time. 3. The apnea test apparatus according to claim 1, wherein the frequency band has a low cut-off frequency higher than a frequency near 0 Hz. 5. The apnea test apparatus according to claim 4, wherein the frequency band is defined by defining a predetermined respiration frequency at a center frequency between the high-frequency cutoff frequency and the low-frequency cutoff frequency. 6. The apnea test apparatus according to claim 5, wherein the high-frequency cutoff frequency is a frequency higher by 0.1 Hz than the predetermined respiratory frequency, and the low-frequency cutoff frequency is a frequency lower by 0.1 Hz than the predetermined respiratory frequency. . 3. The respiration frequency detection unit according to claim 1, wherein the respiration frequency detection unit detects the respiration frequency after frequency analysis of the third pulse wave data every time the apnea determination unit defines the apnea period. An apnea test apparatus characterized by comprising: 3. The data extraction unit according to claim 1, wherein the data extraction unit performs FFT on the third pulse wave data including the apnea period to obtain frequency characteristic data, and on the frequency characteristic data, A respiration frequency detecting means for detecting the respiration frequency; and an inverse FFT means for performing a fast inverse Fourier transform process on the frequency band of the frequency characteristic data to obtain the low frequency data. Respiratory test device. 9. The apnea test apparatus according to claim 8, wherein the respiratory frequency detection means detects a frequency of a spectrum in a range of 0.1 Hz to 0.6 Hz in the frequency characteristic data as the respiratory frequency. 3. The apnea test apparatus according to claim 1, wherein the apnea type determining means includes waveform deformation processing means for exaggerating a pulsation mode of the apnea period data based on the respiratory frequency. . 3. The half cycle of the breathing frequency is ΔT, the apnea period data at a time point t is D _E-S (t), and the apnea period data at a time point (t + ΔT) is D _E-S. Assuming that (t + ΔT), the apnea type determination means is D _E−S (t + ΔT) / D _E−S (t) when D _E−S (t) ≦ D _E−S (t + ΔT) , D _E−S (t)> D _E−S (t + ΔT) when D _E−S (t) / D _E−S (t + ΔT), normalization to generate normalized data in the range 0 to 1 An apnea test apparatus characterized by comprising processing means. 12. The histogram generation means for generating a histogram in which a section obtained by subdividing the range 0 to 1 and the number of data by section of the normalized data belonging to the section correspond to each other, An apnea examination apparatus, comprising: a central degree calculating means for calculating a ratio of the number of data by section belonging to a close section to the total number of data of the normalized data. 13. In claim 12, when the ratio exceeds the first threshold value close to 1, it is determined as central apnea, and whether the ratio is smaller than the second threshold value close to 0 is compared. If the ratio is small, it is determined as obstructive apnea, and the comparison means determines whether the ratio is equal to or lower than the first threshold value and equal to or higher than the second threshold value. An apnea test apparatus characterized by comprising:

Images